Hi Colm,
The FREESURFER NOTE makes reference to both of them.
The a2009s segmentation is brought along for viewing
purposes only (partially as a demonstration of what
one could do, partially because they may be more
detailed), with a potential to be used to generate
probability maps across subjects. That is why I
wrote that they are basically unused.
The aparc+aseg.nii parcellation is used to extract
the white matter and ventricle masks, and is what
was used for the original ANATICOR paper.
It would probably make sense to use 1 parcellation,
you are certainly right about that. But I do not
know the list of 2009s ROIs that would correspond
to those from aparc+aseg.nii (2 for CSF, 10 for WM).
Maybe I could multiply the WM mask from aparc and
multiply it by the 2009a version to get a list,
and maybe compare them with a dice coefficient or
something.
I will make a note of that, thanks!
- rick